Search for a command to run...
One of the most common causes of hospitalization and death among patients with sickle cell disease (SCD) is acute chest syndrome (ACS) [1], defined as a new pulmonary infiltrate on chest radiograph (CXR) in the presence of fever or respiratory symptoms [2]. Causes include both infectious and non-infectious etiologies such as fat embolism [3, 4]. In the healthy population, a CXR is warranted in evaluating a febrile child only when there are respiratory symptoms suggesting pneumonia [5]. Even in neutropenic cancer patients, routine CXR without respiratory symptoms is considered unnecessary [6, 7]. Recommendations for using CXR among children with SCD who present with fever vary. Khan and colleagues [8] and the National Heart, Lung, Blood Institute have recommended against empiric CXRs without respiratory symptoms [9], while UpToDate encourages a low threshold [10]. A prospective study performed over 25 years ago that evaluated the sensitivity of clinical assessment to detect ACS in febrile children with SCD showed that physician impression underestimated the presence of ACS [11]. The low sensitivity in that study was, in part, attributed to the lack of specific signs and symptoms of early ACS. This observation supported the findings of the Cooperative Study of Sickle Cell Disease report of a normal lung exam as the second most common physical exam finding in ACS after rales [3]. In addition, a retrospective study of > 600 ED visits for children with SCD, found that 62% of children with ACS had a normal lung exam [12]. Evolving treatments for SCD over the last two decades, including increased use of hydroxyurea, immunizations, and penicillin prophylaxis, could impact the prevalence and presentation of ACS, warranting a reassessment of earlier findings. Given the significant morbidity and mortality of ACS, compounded by wide variability in practice [8, 12, 13], clinical assessment alone in the absence of standardized guidelines may contribute to missed diagnoses of ACS. Our study objectives were to determine the sensitivity of physician clinical assessment in detecting ACS, and revisit the prevalence, presentation and outcomes of ACS compared to our observations using the same survey from 1999 [11]. We conducted a prospective cross-sectional observational study at two EDs in a single healthcare institution. When caring for a febrile child with SCD, physicians and physician trainees completed our previously used questionnaire [11] adapted to include data on past history of ACS and asthma (Figure S1), given the known association of these risk factors with ACS [12, 14, 15]. Questionnaires were filled out before CXR results were available. Children 0–18 years old with SCD of any genotype and a documented fever ≥ 38.3°C within the past 48 h were eligible. An ACS diagnosis was defined as a new pulmonary infiltrate interpreted by a pediatric radiologist blinded to the study in the presence of fever. Sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and accuracy of physician clinical suspicion of ACS and clinical characteristics were compared between children with and without ACS. The complete methodology and statistical analysis are described in the Supporting Information: Methods. The study protocol was approved by the Institutional Review Boards at Emory University School of Medicine and Children's Healthcare of Atlanta (CHOA). A total of 180 encounters of fever in children with SCD were included, of which 9 episodes were excluded due to the absence of recorded temperature. Of the remaining 171 encounters, 91 (53%) were males, and the mean age was 6.4 ± 5.4 years. ACS was documented in 26 encounters (15.2%) among 20 patients, with 3 patients experiencing 2 episodes. Children with ACS were significantly older than those without ACS (mean age 7.2 ± 4.3 vs. 5.3 ± 4.5 years, p = 0.04). No other significant demographic differences were observed (Table 1). Risk factors significantly associated with a positive CXR included history of ACS (OR 3.53, 95% CI 1.45–8.94, p = 0.006) and chest pain (OR 2.95, 95% CI 1.12–7.31, p = 0.037). Seventy-three percent of ACS episodes had a normal lung exam. ACS was associated with a significantly higher admission rate (OR 43.4, 95% CI 7.52–425.6, p < 0.0001), and BiPAP use (p = 0.02, Table 1). Clinical assessment had a sensitivity of 42.3% and a specificity of 86.2% with a 35.5% PPV and 89.3% NPV and an overall accuracy of 83%. Attending clinical assessment was 37.5% sensitive and 92.3% specific for ACS, with PPV and NPV of 50.0% and 87.8%, respectively. Trainee clinical assessment was 50% sensitive and 79.1% specific for ACS, with PPV and NPV of 26.3% and 91.4%, respectively (Table 2). Of the 26 episodes of ACS, only 11 (42%) were suspected to have ACS by the evaluating physician. The only statistically significant differences between suspected and unsuspected ACS were maximum temperature and abnormal lung physical exam (Table S1). All patients with ACS had at least 1 risk factor based on the CHOA criteria (ACS history, cough, chest pain, abnormal lung exam) [12], whereas 4 (15%) had no risk factors based on NHLBI criteria (shortness of breath, cough, rales, tachypnea) [9] (Table 3). There were no statistically significant differences in hospital course of patients with suspected versus occult ACS (Table S2). Among febrile children with SCD presenting to the ED, 15% were diagnosed with ACS confirmed by CXR. Supporting our prior observation 25 years ago, physician clinical assessment showed poor sensitivity. Although a history of ACS and presence of chest pain were significantly associated with ACS as previously reported [12], more than two-thirds of patients had a normal lung exam, also consistent with previous reports [3, 11]. Nearly, 60% of ACS cases were unsuspected, of which 93% had a normal lung exam, however the clinical course did not differ between suspected and unsuspected cases of ACS. This suggests a similar ACS severity among suspected and occult ACS, where milder presentations did not translate to more favorable outcomes. Patients with suspected ACS had a higher maximum temperature and more frequent abnormal lung examination, suggesting that subtle presentations of ACS may be missed in the absence of CXR assessment. Given the high mortality and morbidity, reliable diagnostic tools or biomarkers may be necessary for more timely diagnosis and management [1, 4, 16-18]. The prevalence of ACS was lower in our study than in 1999 (15% vs. 24%) [11], likely reflecting hydroxyurea use, penicillin prophylaxis, and pneumococcal vaccination [4, 19]. The poor sensitivity of clinician impression for diagnosing ACS is confirmed in this study, echoing our 1999 results [11], where ACS was underestimated by evaluating clinicians. In the 1999 study, clinical assessment alone missed 61% of ACS cases, similar to the 58% of unsuspected ACS in this study. Trainee suspicion of ACS was higher than attending physicians, but specificity was lower, mirroring prior observation [11]. Vital signs and respiratory symptoms remain of limited value in predicting ACS [11, 12]. Current NHLBI guidelines recommend a CXR for febrile children with SCD only if they present with respiratory symptoms, including shortness of breath, tachypnea, cough, or rales [9]. However in our study, we did not find any statistically significant lung exam findings that raised suspicion for ACS. Over 70% of patients with ACS presented to the ED with a normal lung exam, consistent with previous studies [3, 11]. Both a history of ACS and presence of chest pain were statistically significant findings associated with ACS, supporting the documented recurrent nature of ACS [8, 12]. Although one recent retrospective study reported that all patients with ACS showed respiratory symptoms, this was based on a broader criterion that included chest pain [8], currently not part of the NHBLI guidelines [9]. In our prospective study, 15% of ACS episodes had no risk factors according to NHLBI criteria (shortness of breath, cough, rales, tachypnea) [9]. Interestingly, all patients with ACS had at least 1 risk factor based on our previously reported CHOA criteria (history of ACS, cough, chest pain or abnormal lung exam) [12]. Although a history of ACS was not collected in our original survey, these results otherwise echo the findings of the original 1999 prospective study [11] and confirm the observations of the recent retrospective study [8], emphasizing that reliance solely on NHLBI-defined respiratory symptoms may miss early ACS. Incorporating history of ACS or the presence of chest pain into the criteria for obtaining CXR in febrile children with SCD is warranted, even in the presence of a normal lung exam. While some studies have shown an association between history of asthma and risk of ACS [14, 15], asthma history was similar in patients with and without ACS, consistent with a recent retrospective study [8]. However since these studies only evaluate ACS during one point in time, they did not assess the relationship between asthma and any history of ACS, which has been established [14, 15]. Considerable practice variation exists in how clinicians approach fever in SCD [13]. Additional context regarding imaging utilization patterns, radiation considerations, point-of-care lung ultrasound, and emerging biomarkers is provided in the Supporting Information: Discussion. Although the prevalence of ACS has decreased over the last 25 years, it remains common among febrile children with SCD. In this prospective study, over 70% of children with ACS presented with a normal lung exam, consistent with prior reports [3, 11], likely contributing to an underestimation of ACS by physicians. However, these findings should be interpreted considering study limitations (Supporting Information: Discussion). Given the high mortality and morbidity associated with ACS, clinicians should be aware of risk factors and maintain a low threshold to obtain a CXR when evaluating febrile children with SCD. C.A.R. was supported in part by the US National Institutes of Health/NHLBI (K23HL173694). C.R.M. was supported in part by the US National Institutes of Health/National Center for Complementary and Integrative Health (K24 AT009893). The funders had no role in the study design or in the collection, analysis, or interpretation of the data. The funders did not write the report and had no role in the decision to submit the paper for publication. This information or content and conclusions are those of the author and should not be construed as the official position or policy of, nor should any endorsements be inferred by the U.S. Government. This study was approved by the Institutional Review Boards of Emory University School of Medicine and Children's Healthcare of Atlanta. The authors of this manuscript have no competing interests and nothing to declare related to this manuscript. C.R.M. is the inventor or co-inventor of several UCSF-Benioff Children's Hospital Oakland patents that include nutritional supplements and is an inventor/co-inventor of several Emory University School of Medicine patents/patent-pending applications for nutritional supplements for autism, coronaviruses, and pain, is a consultant for CSL Behring, F. Hoffmann-La Roche LTD, Octapharma and Pfizer, is on the Scientific Advisory Board of TRILITY, is an editor the Sickle Cell Disease-Fever and Sickle Cell Disease-Pain reference for UpToDate, is the Founder and Executive Director for Food as Medicine Therapeutics LLC and Sigma Spero LLC, is a member of American Society of Hematology (ASH) Sickle Cell Disease Research Priorities working group since February 2024, and has received research support from the United States Food and Drug Administration, Health Resources & Services Administration (HRSA, agency of the US Department of Health and Human Services) and the National Institutes of Health. The data that support the findings of this study are available from the corresponding author upon reasonable request. Data S1: Supporting Information: Methods. Figure S1: Study questionnaire for physician assessment of acute chest syndrome (ACS) in children with sickle cell disease and fever, including items on vital signs, past medical history, presenting symptoms, physical examination, physician training level, chest radiograph findings, ACS diagnosis, and patient disposition (admission or discharge). Abbreviations: SCD, sickle cell disease; temp, temperature; HR, heart rate; RR, respiratory rate; BP, blood pressure; pulse ox, pulse oximetry; H/O, history of; SX, symptoms; ACS, acute chest syndrome; CXR, chest x-ray; PGY, postgraduate year; PEM, pediatric emergency medicine. Table S1: Comparison of patients with suspected vs. unsuspected ACS. Table S2: Clinical course in suspected versus unsuspected ACS. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.